Research based on organizational learning perspective consider IT innovation diffusion to be the result of the learning process, emphasizing learning within a single organization and biased in favor of learning-by-doing [15] . In the face of complex technology, organization usually formed knowledge barriers due to the learning burden, and will delay the adoption of complex technology until the knowledge barriers were lowered, or until the organization obtained enough skills to successfully implement and use the technology [16] . Fichman and Kemerer (1999) [17] argued that knowledge barrier is one of the reasons that lead to assimilation gap in the

The chart source: according to the relevant literature review.

process of IT innovation, organizations adopt the new information technologies might not be able to deploy them. Therefore, organizational learning is an important perspective for exploring IT innovation diffusion.

Fichman and Kemerer (1997) [18] considered that organizations with greater learning-related scale have a greater opportunity to capture economies of scale in learning, and learning costs are more likely to be diluted. Organizations that have more related knowledge, the knowledge and behavior are more diversity, are more likely to accept or learning new knowledge needed for innovation. Results of empirical study show that learning-related scale, related knowledge and diversity of knowledge have positive impact on the assimilation of object-oriented programming languages (OOPLs) within the organization.

From the perspective of organizational learning based on experience, Sherif et al. (2006) [19] studied the goal conflict problems of software reuse in the process of assimilation. He considered that the organization conduct conceptual and operational learning based on past experience, and form innovation-related knowledge along with the accumulation of experience. Then, the rich knowledge can reduce the level of conflict in the process of IT innovation assimilation. Research results show that organizational learning have negative influence on goal conflict, so as to ensure the assimilation of IT innovation inside the organization, and then bring a series of output of software reuse.

Many scholars (e.g. Carlo et al., 2012 [20] ; Saraf et al., 2013 [21] ; Cooper and Molla, 2014 [22] ) studied IT innovation of enterprises from the perspective of organizational learning based on absorptive capacity. Organization developed absorptive capacity in the process of recognize, absorb, transform and apply valuable external knowledge, and actively affect the assimilation of complexity IT innovation [23] . Carlo et al. (2012) [20] studied the organizational IT innovation behavior from the cognitive and behavioral dimensions of absorptive capacity. Study results show that the more diverse and linkable of existing knowledge, the more perception of innovation-related knowledge from external environment, and organization can improve the level of innovation through learning this knowledge, finally resulting in IT base innovation and IT process innovation.

Social contagion refers to a series of processes that prior adopters of IT innovations cause contagious influence to the organizations which have not yet adopt the IT innovations. When more and more organizations adopt the innovation, nonadopters will feel the social pressure, thus forming the phenomenon of social contagion. Social contagion view explain IT innovation diffusion mainly by neo-institutional theory and social learning theory. Neo-institutional theory is concerned with institutional forces (e.g. coercive force, normative force and mimetic force) that lead firms residing in an organizational field to increasingly resemble each other, resulting in institutional isomorphism. Social learning theory overlaps with the mimetic force in positing that people engage in a process of social learning by examining the actions of similar peers [1] . All prior adopters can be viewed as “carriers” or sources of contagious influence. The role of institutional forces, inter-organizational observation, information transmission and imitation learning made the nonadopters infected with the source of infection [24] .

The social contagion phenomenon in the field of IT innovation has been paid more and more attention by scholars. Teo et al. (2003) [25] first applied the institutional theory to study the adoption of enterprise financial electronic data interchange (FEDI) system. The empirical results show that the coercive force, normative force and mimetic force which exist in external institutional environment positively impact the adoption of the FEDI system within organization. Liang et al. (2007) [26] believed that the institutional pressures not only affect the adoption and implementation of IT innovation in enterprises, but also affect the assimilation of innovation. Take the enterprise resource planning (ERP) system as the research object, the results of the study support their views. Angst et al. (2010) [24] believed that infection is a complex and dynamic process, the study of social contagion must consider three factors of organizational social structure: susceptibility of nonadopters, infectiousness of infestor, physical and social proximity between nonadopters and infection source. The empirical results show that the above factors have significant relationship with social contagion.

In the field of IT, fashion is a common phenomenon [27] . Though IT innovations always have technical component, they are also managerial innovations [28] . Management fashion plays a significant role in the process of IT innovation diffusion.

Management fashion is a relatively transitory collective belief that an administrative technique is new, efficient, and at the forefront of management practice [29] . The basic concept that apply management fashion to IT innovation diffusion field is: technology suppliers, consultants, analysts, journalists, scholars and potential adoption organizations together to form a community, when IT innovation occurs, community members participate in the discussion of innovation with their own motives [3] , which make community full of speeches about IT innovation and form a fashion trend of innovation. Managers are influenced by macro socio-psychological when actively looking for new technology, certain innovations may garner managerial attention and organizational adoption out of all proportion to the ultimate benefits flowing from their actual use [1] [14] .

Wang (2009) [3] made an empirical study on popularity of ERP, he suggested that 1) the popularity of ERP was influenced positively by the prevalence of highlighted business problems that ERP was claimed to solve; 2) ERP’s popularity was influenced negatively by the prevalence of related innovation concepts; and 3) these influences largely disappeared after ERP passed its peak popularity. Baskerville and Myers (2009) [27] argued that the popularity of IT in the community is the result of amplification. Community members may exaggerate the value of IT innovation, urgency of innovation adoption and ability to implement innovation based on their own purpose, lead IT innovation become a kind of fashion at a faster rate and attract more attention of manager, so as to promote the adoption of IT innovations.

Swanson and Ramiller (2004) [12] discussed the innovation behavior of two different types of enterprises under the tide of fashion. In the face of enterprise innovation, mindful firms will make inferences based on their own conditions. Therefore, the mindfulness innovation behaviors of mindful firms are rational, and this type of firms may produce a boycott of the fashion tide. The mindless firms lack consideration of organization specific characteristic and pay little attention to identifying and exploring new IT innovations. Therefore, when the fashion wave of IT innovation formed, mindless firms are easily affected and usually adopt IT innovation without thinking. Hashim et al. (2014) [30] argued that fashion wave easily lead to “me too”, in this case, IT innovation adoption behaviors may be irregular, unnecessary and invalid, which is not conducive to the further diffusion of innovation. Their empirical research results confirmed that fashion effect have positive effect on innovation adoption, but is not conducive to the implementation of IT innovation.

Organizing vision is an important branch of IT innovation diffusion research in recent years [31] . An organizing vision is a focal community idea for the application of information technology in organizations, community members including technology suppliers, consultants, industry experts, prospective adopters and scholars [28] [12] . When IT innovation occurs, community members immediately open up discussion on some issues, such as the uncertainty and prospect of innovation, and how to embed and utilize information technology in organizational structures and processes, thus forming a vision of IT innovation. If management fashion is temporary collective beliefs suspected of exemplifying superstitious learning, then organizing visions go beyond mere fashion statements to synthesize and carry community wisdom that may draw from or support real learning [15] . Organizing vision describes the usefulness and value of IT in a social way, making it easier for policymakers to decide whether to participate in innovation.

Swanson and Ramiller (1997) [28] believed that the organizing vision can promote the adoption and diffusion of IT innovation in organizations through three basic functions: interpretation, legitimation and mobilization. Kaganer et al. (2010) [4] further subdivided legitimacy into four categories: cognitive, pragmatic, normative and regulative legitimacy, provided a more sophisticated framework for the perspective of organizing vision in study of IT innovation diffusion. Wang and Swanson (2007) [32] took professional services automation (PSA) as research object, they confirmed that the institutional entrepreneurs struggled to legitimate PSA by developing a coherent organizing vision that incorporated compelling success stories, raising the possibility of PSA implementation in the organization. Standing et al. (2013) [33] believed that the organizing visions do not necessarily promote the adoption of IT innovation. Companies rely on organizing visions to judge whether IT innovations have tactical or strategic contribution, in this sense, the key point that decide whether IT innovations have value is not the innovation itself, but the value perception of organizing vision. Thus, when the organizing visions fail to properly perceive the value of IT innovation, it may limit the adoption of IT innovative within the organization [33] .

Kling and Iacono (1994) [34] defined the computerization movement as “a kind of movement whose advocates focus on computer-based systems as instruments to bring about a new social order... based upon collaborations of participants with diverse interests”. In other words, the advocates of computerization movement are trying to establish a new social order through the new IT. As a result, IT innovation has been brought into a broader and more complex macro social environment by computerization movement. The core idea of the computerization movement theory is that the diffusion of IT innovation is not only the product of the scientific research laboratory and the enterprise, but also the product of the social movement.

The main character of the computerization movement is the collective actions, including technological action frames and public discourses. Technological action frame is a comprehensive understanding of how technology works and being used, and it is established in the dynamic process of the interaction between the participants. Public discourse is the expression way of technological action frame and participants ideology, it supports the relationship between computerization and the new social order, and be used to legalize or against the computerization input [31] . Collective actions can promote the movement of computerization by shaping public discourse, and the core of “shaping” is “framing”, refers to a process by which social meanings are constructed, disseminated, and stabilized in discourse, namely, the process of the establishment of technological action frames [35] . With the passage of time, in the role of collective actions, the organization may insert beliefs and ideas of computerization movement into practice and adopt IT innovation.